A model-based parameter estimation technique for wideband interpolation of periodic moment method impedance matrices

Abstract

Several computational electromagnetic modeling techniques have been developed specifically for the analysis and design of frequency selective surfaces (FSS). One of the most popular and widely used of these techniques is the periodic method of moments (PMM). The application of the PMM to the analysis of FSS geometries at a single frequency involves the construction and solution of a linear system of equations of the form Zi=v, where Z is the impedance matrix, i is the vector that contains the unknown coefficients of the basis functions used to expand the surface current, and v is the known excitation vector. For simulations where wideband analysis is required, the entire system has to be reconstructed and solved for each frequency sampling point in the band of interest. Therefore, finding a way to reduce the time involved in calculating the Z matrix is extremely important. We introduce a model-based parameter estimation (MBPE) technique that can be used to accelerate the computation of the PMM impedance matrix, Z, over wide frequency ranges for efficient analysis and design of FSS.